Enterprise supplementary medical reimbursement holographic account driven closed-loop management method

Abstract

The application discloses a kind of based on enterprise supplementary medical reimbursement holographic account book driven closed-loop management method, comprising: initialization includes finite state machine instance and holographic account record of hash chain storage area;Each link is started before Hash check, after starting link, output Hash is written into hash chain and forms tamper-evident chain, and process is automatically migrated by state machine driving;Extraction bill fiber texture and papermaking equipment microscopic imprint feature combination is combined as joint physical fingerprint, and business data hash value is bound to generate double-layer fusion verification vector by HMAC-SHA256;The vector is compared with historical record multidimensional and checks hash chain integrity, distinguishes abnormal type and triggers different processing;Abnormality intercepts and triggers review, adopts exponential weighted moving average method adaptive adjustment judging threshold based on review result, forms account driven closed-loop traceability chain;The present application realizes account initiative driving, data hash chain closed-loop locking, bill physical feature double-layer source and detection parameter adaptive optimization.

Description

Technical Field

[0001] This invention relates to the field of data processing and process management technology, specifically to a management method for enterprise supplementary medical insurance reimbursement, and more particularly to a full-process management method for enterprise supplementary medical insurance reimbursement based on holographic ledger finite state machine driving, link hash chain closed-loop locking, paper physical characteristics dual-layer traceability and adaptive feedback closed loop. Background Technology

[0002] Corporate supplementary medical insurance is a supplementary medical security system established by enterprises for their employees on the basis of participation in national basic medical insurance. The reimbursement process of supplementary medical insurance involves multiple business links, such as invoice submission, scanning and collection, OCR information recognition, amount verification, invoice authenticity verification, rule review, payment execution, and archiving. Data interaction between each link is frequent, and the review rules are complex, which places high demands on the standardization of process management, data consistency, and risk control capabilities.

[0003] Based on a systematic analysis of existing technologies, the current management of supplementary medical insurance reimbursement for enterprises has the following shortcomings: First, process control relies on manual triggering or external rule engines, lacking a proactive, automated closed-loop mechanism centered on ledgers. In existing technologies, the advancement of each stage of the reimbursement process typically depends on manual triggering by finance personnel or the activation of a pre-set rule engine based on a single condition, with the ledger merely serving as a passive data recording container. When multiple applications occur concurrently or anomalies arise, the system cannot autonomously halt or trace the process, leading to extended processing cycles and low efficiency in corrective actions.

[0004] Second, the data at each stage lacks closed-loop verification mechanisms to prevent tampering. In the reimbursement process, critical business data flows sequentially between multiple stages, posing a risk of inconsistency due to system errors, transmission failures, or human tampering. Existing technologies typically record data independently at a single stage, lacking a verification mechanism to link the input and output data of all stages into an immutable chain of evidence. This makes it difficult to accurately pinpoint the source of anomalies.

[0005] Third, the anti-counterfeiting verification methods for invoices are simplistic and ineffective in addressing the risk of invoice reuse. Existing technologies primarily rely on repeated comparisons of structured information such as invoice codes and numbers, or on single-dimensional physical feature comparisons using paper fiber texture. When different invoices come from the same paper batch, the statistical similarity of paper textures is high, making accurate differentiation difficult using only single-layer physical features. Furthermore, there is a lack of inseparable binding verification between physical features and business data, allowing tampering in any dimension to bypass detection.

[0006] Fourth, there is a lack of an adaptive closed-loop learning mechanism based on review feedback after anomaly detection. Existing technologies only output alarms or block after detecting anomalies, and the results of manual review are not systematically fed back to the detection model to optimize the judgment parameters, making it difficult for false and missed interception rates to continuously improve with business accumulation. Summary of the Invention

[0007] This invention aims to provide a closed-loop management method driven by a holographic ledger for enterprise supplementary medical reimbursement, in order to solve the technical problems in the existing technology, such as insufficient degree of process automation and closed loop, weak data consistency verification methods, single dimension of invoice anti-counterfeiting verification, and lack of adaptive feedback mechanism. It realizes ledger proactive driving, full-process data closed-loop locking, dual-layer traceability of paper physical characteristics, and adaptive optimization of detection parameters.

[0008] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: Step S1: Holographic ledger initialization and state machine instantiation.

[0009] The system receives medical reimbursement invoices submitted by the applicant, generates a globally unique batch number for each reimbursement application, and initializes a holographic ledger record uniquely bound to that batch number. The holographic ledger record includes a master ledger table, a hash chain storage area for ledger processes, and a ledger papermaking equipment imprint storage area.

[0010] The master ledger table must include at least the batch number, applicant identifier, process status field, expected hash values ​​of input data for each stage, and hash values ​​of output data for each stage. The process status field contains a finite state machine instance, which predefines a finite set of states and a set of state transition rules for the entire supplementary medical insurance reimbursement process. The state transition rule set defines the legal transition paths between states and the transition triggering conditions. Transition triggering conditions include at least one of the following: hash verification pass signal for the current stage, processing completion signal for the current stage, and physical traceability verification pass signal.

[0011] The ledger hash chain storage area stores hash records for each stage in append-only mode. Each hash record contains the current stage identifier, the hash value of the current stage's output data, and the hash value of the previous stage's output data. The hash values ​​of each stage are linked in the order of stage flow to form an immutable hash evidence chain. The ledger papermaking equipment imprint storage area stores the periodic microscopic imprint features left on the paper by the papermaking equipment during the papermaking process.

[0012] Step S2: Ledger-driven hash verification and state transition at each stage.

[0013] Before each step of the reimbursement application process begins, the hash value of the output data from the previous step, recorded in the master ledger, is used as the expected hash value for the input data of the current step. The actual input data for the current step is then hash-verified. Only after successful verification is the current step allowed to proceed.

[0014] After the current stage is completed, the hash value of the output data generated in the current stage is written to the main ledger table, and simultaneously written to the hash chain storage area of ​​the ledger stage and associated with the batch number, so that the hash values ​​of each stage are linked in the order of stage flow to form an immutable hash evidence chain. At the same time, the processing result and operation timestamp of the current stage are written to the main ledger table, and the finite state machine instance automatically drives the process state to the next state according to the state transition rule set.

[0015] The hash chain storage area for the ledger process adopts an append-only write mode. Each newly written hash value contains a link reference to the hash value of the previous stage. Based on the collision resistance and avalanche effect of cryptographic hash functions, any tampering with the data in any stage of the hash evidence chain will cause a significant change in the output hash value of that stage, thereby causing the hash verification of the next stage to fail. The system can thus accurately detect tampering and trace back along the hash chain to locate the specific stage where the anomaly was introduced.

[0016] Step S3: Physical source extraction and two-layer fusion verification vector generation.

[0017] Optical acquisition is performed on medical reimbursement receipts to obtain images of the receipts. Optical acquisition can be accomplished through at least one of the following methods: a mobile terminal equipped with a macro lens, a high-resolution flatbed scanner, or a multispectral imaging device. The acquisition resolution should be sufficient to clearly capture the microscopic texture details of the paper.

[0018] First texture data reflecting the inherent microscopic characteristics of the physical medium of the paper is extracted from the image of the paper document, and a first fingerprint is generated based on the first texture data. The first texture data includes at least the fiber distribution texture data of the paper document, which is an individual-level fingerprint of the physical characteristics of the paper document.

[0019] Simultaneously, the periodic microscopic imprint features left by papermaking equipment during the paper production process are extracted from the document image to generate a second fingerprint. These periodic microscopic imprint features left by the papermaking equipment include at least one of the following: microscopic indentation patterns of the warp and weft threads left on the paper surface by the forming wire of the paper machine; texture imprints transferred to the paper surface by the press blanket through mechanical pressure; or surface microscopic smoothness distribution features formed by the influence of the microscopic roughness of the drying cylinder surface during the paper drying process. The second fingerprint belongs to the batch-level fingerprint of the document's physical characteristics.

[0020] By combining the first and second fingerprints into a joint physical fingerprint, a two-tiered traceability system based on the physical characteristics of the invoices can be achieved at both the individual and batch levels.

[0021] Extract structured business data from invoices, including fields such as invoice code, invoice number, invoice date, and amount, and calculate the cryptographic hash value of the structured business data. Cryptographically bind the joint physical fingerprint to the cryptographic hash value, using the joint physical fingerprint as the message and the cryptographic hash value as the key, and employ the HMAC-SHA256 algorithm to generate a two-layer fusion verification vector that contains both physical traceability features and business features that are inseparable.

[0022] Step S4: Multidimensional matching and comparison and anomaly type determination.

[0023] The dual-layer fusion verification vector is compared with the dual-layer fusion verification vector of historical reimbursement documents pre-stored in the holographic ledger archive database in a multi-dimensional matching process, while verifying the integrity of the hash evidence chain in the hash chain storage area of ​​the ledger process.

[0024] Multidimensional matching and comparison includes joint physical fingerprint similarity calculation in the physical tracing dimension and hash consistency comparison in the business dimension. The physical tracing similarity calculation is performed on the first and second fingerprint dimensions respectively, and then the two dimensions' similarities are weighted and fused to obtain the comprehensive physical tracing similarity. The business dimension hash consistency comparison employs strict consistency determination.

[0025] The hash chain integrity verification in the ledger process includes: verifying whether the output data hash value of each stage is consistent with the expected input data hash value of the next stage according to the stage flow order, and verifying whether the link references of the hash values ​​of each stage are continuous and have not been tampered with.

[0026] Based on the results of multi-dimensional matching and comparison, the abnormal types corresponding to reimbursement applications are distinguished. Abnormal types include at least: normal types when both physical traceability similarity and business consistency exceed their respective preset thresholds; suspected invoice reuse types when physical traceability similarity exceeds a preset threshold but business consistency is below a preset threshold; and process data tampering types when the hash chain integrity verification in the ledger stage fails. Differentiated processing strategies are triggered based on the abnormal type.

[0027] Step S5: Exception handling and adaptive closed-loop management.

[0028] When the multidimensional matching comparison indicates an anomaly, an interception record is generated and a review process is triggered. The review result label is then fed back to the holographic ledger record. Based on the review result label, the similarity threshold for the physical traceability dimension and the consistency threshold for the business dimension are adaptively adjusted. An exponentially weighted moving average method is used to dynamically determine the adjustment step size based on the confidence level of the review result. Simultaneously, the anomaly type, interception reason, review result, and adjusted threshold are written into the holographic ledger record, forming a ledger-driven closed-loop management traceability chain.

[0029] The logical relationship between the steps in this invention is as follows: Steps S1 to S5 are executed sequentially in chronological order; Step S2 is triggered at each business processing stage of the reimbursement application, including pre-start verification and writing two sub-steps with sequential dependencies after completion; Step S3 is executed when the process enters the physical traceability verification state, and each sub-step is completed in chronological order; Step S4 is executed after the two-layer fusion verification vector is generated in Step S3; Step S5 is executed after Step S4 determines that an anomaly has occurred.

[0030] In addition, when any enterprise node receives a reimbursement application, it can homomorphically encrypt the two-layer fusion verification vector and broadcast it to other enterprise nodes participating in the joint defense through the consortium blockchain network. Each participating enterprise node performs ciphertext field matching on the received encrypted vector locally, outputs the ciphertext field verification result and returns it to the requesting node. The requesting node decrypts and weights the verification results returned by each node, comprehensively judges whether there is cross-enterprise duplicate reimbursement, and writes the collaborative verification result into the holographic ledger record.

[0031] Compared with the prior art, the beneficial effects of the present invention are as follows: First, it achieves automated closed-loop management of the entire reimbursement process driven by the ledger. By embedding a finite state machine instance into the ledger master table, the control logic for process state transitions is built into the ledger. The ledger is upgraded from a passive recording container to an active driving engine. After each step is completed, the state machine automatically determines the transition conditions and drives the state transition, without the need for manual or external rule intervention.

[0032] Second, an immutable end-to-end hash evidence chain has been constructed. The hash values ​​of the output data at each stage are linked in the order of flow to form a chain structure. Before the next stage starts, the consistency between the input data and the upstream output data is automatically verified. If any data is tampered with at any stage, the hash verification will fail and the link will be broken. The system can trace back along the chain to accurately locate the stage where the anomaly was introduced, upgrading data integrity protection from single-point defense to full-link closed-loop defense.

[0033] Third, it achieves two-tiered traceability of physical characteristics of invoices at both the individual and batch levels. By combining paper fiber texture features with papermaking equipment imprint features into a joint physical fingerprint, the former utilizes the randomness of fiber distribution for individual differentiation, while the latter leverages the batch similarity of equipment imprints for batch verification, solving the problem of difficulty in distinguishing papers from the same batch. The cryptographic binding of the joint physical fingerprint with the business data hash value makes physical verification inseparable from business verification, effectively preventing single-dimensional bypass attacks.

[0034] Fourth, an adaptive closed-loop learning mechanism based on review feedback was established. The results of manual review are fed back to the ledger, and the judgment threshold is adjusted adaptively accordingly. As business data accumulates, the system continuously optimizes the detection parameters, gradually reducing the false interception rate and the missed interception rate.

[0035] Fifth, it achieves multi-dimensional, end-to-end closed-loop management and traceability. The holographic ledger fully records the entire lifecycle data from invoice submission to threshold adjustment, including process status migration paths, hash verification records, comparison results, review results, and parameter adjustment history, all of which are protected by hash evidence chains, providing comprehensive and tamper-proof data support for enterprise auditing and compliance inspections. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0037] The present invention will be further described in detail below with reference to specific embodiments. It should be noted that the following embodiments are only used to illustrate the technical solutions of the present invention and do not constitute a limitation on the scope of protection of the present invention. Specific numerical values, parameters, and exemplary thresholds appearing in the embodiments are for illustrative purposes only, and those skilled in the art can adjust them according to actual application scenarios. Other implementation schemes that can be obtained by those skilled in the art based on the embodiments of the present invention without creative effort are all within the scope of protection of the present invention.

[0038] Example 1: Full Process of Normal Reimbursement Application This example illustrates the complete process of an employee submitting a supplementary medical insurance reimbursement application. Figure 1 The embodiments of the present invention are illustrated below.

[0039] Step S1: Holographic ledger initialization.

[0040] After receiving the medical reimbursement invoices submitted by employees, the enterprise supplementary medical insurance reimbursement management system generates a globally unique batch number for the current reimbursement application and initializes a holographic ledger record uniquely bound to that batch number.

[0041] After initializing the master ledger table, the batch number field is written with the generated unique identifier, the applicant identifier field is written with the employee's unique identity identifier, and the submission timestamp field is written with the current system time. The process status field loads a finite state machine instance, which contains a predefined set of finite states and a set of state transition rules. The finite state set includes nine states: document submission, scan acquisition, information recognition, amount verification, invoice authenticity verification, physical traceability verification, rule review, payment execution, and archiving completion. The state transition rule set defines the legal transition paths and triggering conditions between each state. For example, the triggering conditions for transitioning from the document submission state to the scan acquisition state are the document submission completion signal and the batch number allocation completion signal; the triggering conditions for transitioning from the scan acquisition state to the information recognition state are the scan acquisition completion signal, the document image being written to the file server signal, and the hash verification passing signal.

[0042] The hash chain storage area and the papermaking equipment imprint storage area in the ledger process are initialized to empty.

[0043] Step S2: Ledger-driven hash verification and state transition.

[0044] During the document submission phase, the document image uploaded by the employee is stored on the file server. The system calculates the cryptographic hash value of the original document image file and writes this hash value into the output data hash value field of the ledger master table. Simultaneously, the first hash record is appended to the hash chain storage area of ​​the ledger phase. This record contains the calculated hash value for the current phase's output, while the hash value for the previous phase's output is empty. The finite state machine instance detects the document submission completion signal and the batch number allocation completion signal, automatically transitioning the process state to the scanning and acquisition state.

[0045] Before the scanning and acquisition phase begins, the system reads the output data hash value from the document submission phase from the master ledger table, using it as the expected input hash value for this phase. Simultaneously, it calculates the hash value for the document image to be processed and compares the two. Since the document image has not been modified during the transfer process, the comparison results are consistent, the hash verification passes, and the system allows the scanning and acquisition phase to begin. After image preprocessing is completed, the scanning and acquisition phase calculates the hash value of the standard scanned image and writes it to the master ledger table. A second hash record is appended to the hash chain storage area of ​​the ledger phase, where the output hash value of the previous phase references the output hash value of the document submission phase. The finite state machine instance automatically transitions the process state to the information recognition state.

[0046] The information identification, amount verification, and invoice authenticity verification processes are executed sequentially using the same pattern: a hash verification is performed before initiation; once the verification passes, the process begins; after processing, the hash value of the output data is calculated and written to the ledger master table and the hash chain storage area, with a finite state machine automatically driving the state transition. Through these processes, a continuous hash evidence chain from invoice submission to invoice verification is formed in the ledger's hash chain storage area.

[0047] Step S3: Physical source extraction and two-layer fusion verification vector generation.

[0048] The process enters the physical traceability verification stage. The system performs optical acquisition on the invoice to obtain high-resolution images of the invoice.

[0049] The blank areas of the paper without printing ink are selected from the image of the ticket, and the fiber distribution texture data is extracted as the first texture data. The fiber texture features such as direction, density, and intersection distribution are encoded into feature vectors using a texture analysis algorithm. Then, a perceptual hash algorithm is used to generate the first fingerprint, which is a fixed-length binary sequence that reflects the unique fiber texture features of the ticket paper.

[0050] Simultaneously, a two-dimensional Fourier transform is performed on the same blank area to detect periodic signals in the frequency domain. Periodic peaks corresponding to the warp and weft weaving structure of the paper machine forming wire are identified from the frequency domain, and their frequency, amplitude, and direction parameters are extracted and encoded as a second fingerprint. The second fingerprint reflects the equipment marking characteristics of the paper batch to which the document belongs.

[0051] By concatenating and splicing the first and second fingerprints into a joint physical fingerprint, a two-layer physical traceability system at the individual and batch levels can be achieved.

[0052] Structured business data extracted from OCR during the information recognition stage, including fields such as invoice code, invoice number, invoice date, and amount, is concatenated according to a fixed format and then used to calculate a cryptographic hash value. Using the joint physical fingerprint as the message and the cryptographic hash value as the key, the HMAC-SHA256 algorithm is employed for key derivation and binding, generating a two-layer fusion verification vector. This vector simultaneously contains physical traceability features and business features, and the two are cryptographically inseparable.

[0053] The system will independently store the physical fingerprint in the ledger papermaking equipment imprint storage area, store the cryptographic hash value in the ledger master table, and write the two-layer fusion verification vector into the ledger master table. A finite state machine instance will transition the process state to the rule review state.

[0054] Step S4: Multidimensional matching and comparison and anomaly type determination.

[0055] The system reads the current joint physical fingerprint from the ledger papermaking equipment imprint storage area and retrieves the joint physical fingerprints of historical reimbursement documents from the holographic ledger archive database, calculating the similarity of each physical traceability dimension. The first fingerprint dimension uses Hamming distance as a metric, and after normalization, a first fingerprint similarity score is obtained; the second fingerprint dimension uses cosine similarity as a metric, obtaining a second fingerprint similarity score. The scores of the two dimensions are then weighted and fused according to preset weights to obtain a comprehensive physical traceability similarity. Simultaneously, a strict consistency comparison is performed between the hash value of the current business data and the hash values ​​of historical business data.

[0056] The system also verifies the integrity of the hash evidence chain in the hash chain storage area of ​​the ledger link according to the flow order of the links, verifies whether the output hash of each link is consistent with the expected input hash of the next link, and whether the link references are continuous and unbroken.

[0057] Upon comparison, the overall physical similarity of the reimbursement application exceeded the preset threshold, but the hash consistency comparison result at the business dimension was inconsistent, and the hash evidence chain integrity verification passed. This pattern meets the criteria for suspected invoice reuse—indicating that the submitted invoice is highly similar to a previously reimbursed invoice in physical characteristics, but the business data is different, suggesting that the same invoice may have been used for different reimbursement applications. The system classifies this reimbursement application as suspected invoice reuse.

[0058] Step S5: Exception handling and adaptive closed-loop management.

[0059] The system generates an interception record, which includes information such as batch number, anomaly type, interception timestamp, physical traceability similarity score, matched historical batch number, and business dimension comparison results. The interception record is written into the holographic ledger record and a manual review process is triggered.

[0060] After reviewing the images and business information of the current and historical invoices, the manual reviewer found that although the two invoices were highly similar in physical characteristics, there were subtle differences in fiber texture, confirming that they came from different papers from the same invoice album. The reviewer made a "confirmed normal" review conclusion and fed it back to the system.

[0061] After receiving the verification result label, the system adaptively adjusts the comprehensive physical source similarity threshold using the exponentially weighted moving average method. Since this was a system misjudgment, a new threshold is calculated using the relaxed threshold formula, with the adjustment step size determined based on the confidence level. The adjusted threshold is moderately relaxed compared to the original threshold.

[0062] The system records the anomaly type, the reason for interception, the review result, the thresholds before and after adjustment, and the reason for adjustment in a holographic ledger, forming a complete closed-loop management traceability chain. Since the review result confirms that everything is normal, the system releases the reimbursement application for subsequent rule review, payment execution, and archiving.

[0063] Example 2: Process Data Tampering Detection and Anomaly Location This example illustrates the ability of the hash chain in the ledger process to detect and locate data tampering.

[0064] A reimbursement application was successfully completed during the amount verification stage. The hash value of the amount verification result was written to the main ledger table and the hash chain storage area, and the process status had been moved to the invoice authenticity verification stage. Before the invoice authenticity verification stage was initiated, the amount verification result data in the main ledger table was maliciously tampered with.

[0065] Before the invoice authenticity verification process begins, the system reads the hash value of the output data from the amount verification process from the master ledger table as the expected hash value for the input data. It then calculates the hash value for the actual input data and compares it. Because the data has been tampered with, the two hash values ​​are inconsistent, and the hash verification fails. The system prevents the invoice authenticity verification process from starting and generates a hash verification anomaly alarm.

[0066] The system traces back along the hash chain of the ledger process: it checks the integrity of the records in the hash chain storage area for the amount verification step, confirming the existence of the records; it compares the output hash value in the records with the hash value of the amount verification step in the main ledger table; it checks the records for the invoice authenticity verification step, confirming that they have not yet been generated; it traces the actual processing result data, discovering that it has been tampered with. The system identifies the abnormal entry point as the amount verification step, writes the location result into the holographic ledger record, triggers manual review and auditing processes, and promptly prevents further impact from the tampering.

[0067] Example 3: Cross-enterprise joint defense verification This example illustrates the implementation of cross-enterprise joint defense.

[0068] Within a certain enterprise group, the group's financial shared service center, subsidiary A, subsidiary B, and a cooperating insurance institution jointly built a consortium blockchain network, with each participating node deploying blockchain nodes and federated learning clients.

[0069] After an employee of subsidiary A submits an expense reimbursement application, the system generates a two-layer fusion verification vector. While performing multi-dimensional matching and comparison locally, the fusion vector is encrypted using a homomorphic encryption algorithm and then broadcast to other participating nodes through the consortium blockchain network.

[0070] After receiving the encrypted vector locally, Subsidiary B calculates the encrypted distance metric between the encrypted vector and its local historical records in the encrypted domain. The encrypted domain matching result is returned to Subsidiary A. Subsidiary A decrypts the results returned by each node and performs a weighted fusion, with the weight of each node dynamically determined based on its historical verification accuracy. The fusion result shows that Subsidiary B has a highly matching historical record with a payment status. The system determines that there is cross-enterprise duplicate reimbursement, generates a joint prevention and early warning signal, and writes it into the holographic ledger record. Throughout the entire process, the original data of each node does not leave its local domain, achieving cross-enterprise collaborative joint prevention under privacy protection.

[0071] The method described in this invention can be implemented based on existing enterprise information management systems, conventional optical acquisition equipment, and standard cryptographic libraries. Those skilled in the art can implement it without creative effort based on the content disclosed in the specification, and it has good industrial applicability.

[0072] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A closed-loop management method driven by a comprehensive ledger for enterprise supplementary medical reimbursement, characterized in that: Includes the following steps: Step S1: Receive the medical reimbursement invoices submitted by the reimbursement applicant, generate a globally unique batch number for the current reimbursement application, and initialize a holographic ledger record uniquely bound to the batch number; the holographic ledger record includes a ledger master table, a ledger process hash chain storage area, and a ledger papermaking equipment imprint storage area; the ledger master table at least includes a batch number, applicant identifier, process status field, expected hash value of input data for each process, and hash value of output data for each process; the process status field in the ledger master table includes a finite state machine instance, and the finite state machine instance predefines a finite state set and a state transition rule set for the entire supplementary medical insurance reimbursement process; Step S2: Before each business processing step of the reimbursement application is initiated, the hash value of the output data of the previous step recorded in the master ledger table is used as the expected hash value of the input data of the current step. The actual input data of the current step is hash verified. The current step is allowed to start only after the verification is passed. After the current step is completed, the hash value of the output data generated by the current step is written to the master ledger table, and simultaneously written to the ledger step hash chain storage area and associated with the batch number. The hash values ​​of each step in the ledger step hash chain storage area are linked in the step flow order to form an immutable hash evidence chain. At the same time, the processing result and operation timestamp of the current step are written to the master ledger table. The finite state machine instance automatically drives the process state to the next state according to the state transition rule set. Step S3: Optically acquire the medical reimbursement receipt to obtain a receipt image; extract first texture data reflecting the inherent microscopic characteristics of the physical medium of the receipt paper from the receipt image, and generate a first fingerprint based on the first texture data; the first texture data includes at least the fiber distribution texture data of the receipt paper; simultaneously, extract the periodic microscopic imprint features left by the papermaking equipment during the papermaking process from the receipt image to generate a second fingerprint; combine the first fingerprint and the second fingerprint to form a joint physical fingerprint; extract the structured business data of the receipt and calculate the cryptographic hash value of the structured business data; cryptographically bind the joint physical fingerprint and the cryptographic hash value to generate a two-layer fusion verification vector containing physical traceability features and business features; Step S4: Perform multi-dimensional matching and comparison between the dual-layer fusion verification vector and the dual-layer fusion verification vector of historical reimbursement documents pre-stored in the holographic ledger archive database, and simultaneously verify the integrity of the hash evidence chain in the hash chain storage area of ​​the ledger process; the multi-dimensional matching and comparison includes joint physical fingerprint similarity calculation in the physical traceability dimension and hash consistency comparison in the business dimension; based on the results of the multi-dimensional matching and comparison, distinguish the anomaly type corresponding to the reimbursement application, and trigger differentiated processing strategies according to the anomaly type; Step S5: When the result of the multidimensional matching comparison indicates an anomaly, an interception record is generated and a review process is triggered; the review result label is fed back to the holographic ledger record, and the similarity threshold of the physical traceability dimension and / or the consistency threshold of the business dimension are adaptively adjusted based on the review result label. The anomaly type, interception reason, review result and adjusted threshold are written into the holographic ledger record to form a ledger-driven closed-loop management traceability chain.

2. The closed-loop management method based on a holographic ledger for enterprise supplementary medical reimbursement as described in claim 1, characterized in that, In step S1, the finite state set includes at least: invoice submission state, scanning and acquisition state, information identification state, amount verification state, invoice authenticity verification state, physical traceability verification state, rule review state, payment execution state, and archiving completion state; the state transition rule set defines the legal transition paths and transition triggering conditions between each state, and the transition triggering conditions include: hash verification pass signal of the current stage, processing completion signal of the current stage, and / or physical traceability verification pass signal.

3. The closed-loop management method based on a holographic ledger for enterprise supplementary medical reimbursement as described in claim 1, characterized in that, In step S2, the hash chain storage area of ​​the ledger links adopts an append-write mode. Each newly written link hash value contains a link reference pointing to the hash value of the previous link. When the hash verification of any link fails, the finite state machine instance prevents the process state transition, generates a hash verification exception alarm, and traces back along the hash evidence chain to locate the exception introduction node, and writes the location result into the holographic ledger record.

4. The closed-loop management method based on a holographic ledger for enterprise supplementary medical reimbursement as described in claim 1, characterized in that, In step S3, the periodic micro-imprint features left by the papermaking equipment include at least one of the following: the micro-indentation pattern of the warp and weft weave structure left by the forming wire of the papermaking machine on the paper surface, the texture imprint of the press section blanket transmitted to the paper surface by mechanical pressure, or the surface micro-smoothness distribution features formed by the influence of the micro-roughness of the drying cylinder surface during the paper drying process.

5. The closed-loop management method based on a holographic ledger for enterprise supplementary medical reimbursement as described in claim 1, characterized in that, In step S3, optical acquisition of the medical reimbursement receipt includes one or a combination of the following methods: photographing the receipt using a mobile terminal equipped with a macro lens; scanning the receipt using a high-resolution flatbed scanner; or acquiring multi-band data of the receipt's preset anti-counterfeiting area using a multispectral imaging device.

6. The closed-loop management method based on a holographic ledger for enterprise supplementary medical reimbursement as described in claim 1, characterized in that, In step S4, the abnormal types include at least: a normal type when the joint physical fingerprint similarity of the physical tracing dimension and the hash consistency of the business dimension both exceed their respective preset thresholds; a suspected ticket reuse type when the similarity of the physical tracing dimension exceeds the preset threshold but the hash consistency of the business dimension is lower than the preset threshold; and a process data tampering type when the hash evidence chain integrity verification in the hash chain storage area of ​​the ledger link fails.

7. The closed-loop management method based on a holographic ledger for enterprise supplementary medical reimbursement as described in claim 1, characterized in that, The step S4 of verifying the integrity of the hash evidence chain in the hash chain storage area of ​​the ledger link specifically includes: verifying whether the output data hash value of each link in the hash evidence chain is consistent with the expected hash value of the input data of the next link in the sequential flow of the links, and verifying whether the link reference of the hash values ​​of each link in the hash evidence chain is continuous and has not been tampered with.

8. The closed-loop management method based on the enterprise supplementary medical reimbursement holographic ledger as described in claim 3, characterized in that, After tracing back along the hash evidence chain to locate the abnormal node, the process further includes: counting the historical abnormal frequency of each business process, and making differentiated adaptive adjustments to the judgment thresholds corresponding to each process in step S5 based on the abnormal frequency; automatically tightening the judgment thresholds for processes with abnormal frequencies exceeding the preset thresholds, and maintaining or appropriately relaxing the judgment thresholds for processes with abnormal frequencies below the preset thresholds; and writing the results of the differentiated adaptive adjustments into the holographic ledger record.

9. The closed-loop management method based on a holographic ledger for enterprise supplementary medical reimbursement as described in claim 1, characterized in that, Also includes: When any enterprise node receives a reimbursement request, it will homomorphically encrypt the two-layer fusion verification vector and broadcast it to other enterprise nodes participating in the joint defense through the consortium blockchain network. Each participating enterprise node performs ciphertext field matching on the received encrypted two-layer fusion verification vector locally, outputs the ciphertext field verification result, and returns it to the requesting node; the requesting node decrypts and weights the verification results returned by each enterprise node, comprehensively judges whether there is cross-enterprise duplicate reimbursement, and writes the collaborative verification result into the holographic ledger record.

10. A closed-loop management system driven by a comprehensive ledger for enterprise supplementary medical reimbursement, characterized in that, include: An optical acquisition device is used to optically acquire images of medical reimbursement receipts. A processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method as claimed in any one of claims 1 to 9.

Images